arXiv subjects

Sugandha Sharma

Publications and source records attributed to Sugandha Sharma.

7 recordsLinked to original sources

arXiv2026

SiMOS quantum-dot spin qubits enabled by extreme-ultraviolet lithography

The realization of large-scale silicon quantum processors requires spin qubits compatible with advanced semiconductor manufacturing technologies, demanding lithographic processes that combine nanometer-scale precision with exceptional uniformity. Although the highest-performing silicon spin qubits demonstrated to date have relied on electron-beam (e-beam) lithography, its serial exposure process limits reproducibility studies and wafer-scale fabrication. Here, we demonstrate high-performance silicon metal-oxide-semiconductor (SiMOS) spin qubits fabricated using extreme-ultraviolet (EUV) lithography in a 300 mm semiconductor pilot line. We report wafer-scale quantum-dot uniformity metrics, including 100 % room-temperature gate-to-gate leakage yield and sub-nanometer control of critical gate dimensions. We characterize four double-dot systems realized in two triple-quantum-dot devices. Gate set tomography (GST) reveals consistently high fidelities across all four systems, with values up to 99.8 % for SPAM, 99.9 % for single-qubit gates, and 99.1 % for two-qubit gates. The devices exhibit highly reproducible exchange turn-on characteristics of 10-13 dec/V, indicating high fabrication uniformity enabled by EUV patterning. These results establish EUV lithography as a viable manufacturing technology for quantum processors based on high-fidelity SiMOS spin qubits.

cond-mat.mes-hall↗

arXiv2026

Understanding oxide-thickness-dependent variability in dense Si-MOS quantum dot arrays

Achieving uniform and scalable control of semiconductor spin qubits remains a key challenge for large scale quantum computing. In this work, we investigate how gate oxide thickness influences uniformity in dense two dimensional silicon quantum dot arrays. Using a 7 x 7 array fabricated in a 300 mm CMOS-process patterned by EUV lithography, we statistically characterize 392 quantum dots across four different oxide thicknesses. The threshold voltages, capacitances, lever arms, and charging energies are extracted using parallel row based measurements and we identify an optimal SiO2 thickness of 17 nm that minimizes threshold voltage variability below 63 mV standard deviation. Our observations illustrate how multiple sources of disorder can introduce competing oxide-thickness dependencies, resulting in non-monotonic trends. These results provide key design guidelines for dense, scalable silicon spin qubit architectures.

quant-ph↗

arXiv2026

VoiceSHIELD-Small: Real-Time Malicious Speech Detection and Transcription

Voice interfaces are quickly becoming a common way for people to interact with AI systems. This also brings new security risks, such as prompt injection, social engineering, and harmful voice commands. Traditional security methods rely on converting speech to text and then filtering that text, which introduces delays and can ignore important audio cues. This paper introduces VoiceSHIELD-Small, a lightweight model that works in real time. It can transcribe speech and detect whether it is safe or harmful, all in one step. Built on OpenAI's Whisper-small encoder, VoiceSHIELD adds a mean-pooling layer and a simple classification head. It takes just 90-120 milliseconds to classify audio on mid-tier GPUs, while transcription happens at the same time. Tested on a balanced set of 947 audio clips, the model achieved 99.16 percent accuracy and an F1 score of 0.9865. At the default setting, it missed 2.33 percent of harmful inputs. Cross-validation showed consistent performance (F1 standard deviation = 0.0026). The paper also covers the model's design, training data, performance trade-offs, and responsible use guidelines. VoiceSHIELD is released under the MIT license to encourage further research and adoption in voice AI security.

cs.SD↗

arXiv2025

Precision high-speed quantum logic with holes on a natural silicon foundry platform

Silicon spin qubits in gate-defined quantum dots leverage established semiconductor infrastructure and offer a scalable path toward transformative quantum technologies. Holes spins in silicon offer compact all-electrical control, whilst retaining all the salient features of a quantum dot qubit architecture. However, silicon hole spin qubits are not as advanced as electrons, due to increased susceptibility to disorder and more complex spin physics. Here we demonstrate single-qubit gate fidelities up to 99.8% and a two-qubit gate quality factor of 240, indicating a physical fidelity limit of 99.7%. These results represent the highest performance reported in natural silicon to date, made possible by fast qubit control, exchange pulsing, and industrial-grade fabrication. Notably, we achieve these results in a near-identical device as used for highly reproducible, high-fidelity electron spin qubits. With isotopic purification and device-level optimisations in the future, our hole spin qubits are poised to unlock a new operation regime for quantum CMOS architectures.

cond-mat.mes-hall↗

arXiv2024

Toward Human-AI Alignment in Large-Scale Multi-Player Games

Achieving human-AI alignment in complex multi-agent games is crucial for creating trustworthy AI agents that enhance gameplay. We propose a method to evaluate this alignment using an interpretable task-sets framework, focusing on high-level behavioral tasks instead of low-level policies. Our approach has three components. First, we analyze extensive human gameplay data from Xbox's Bleeding Edge (100K+ games), uncovering behavioral patterns in a complex task space. This task space serves as a basis set for a behavior manifold capturing interpretable axes: fight-flight, explore-exploit, and solo-multi-agent. Second, we train an AI agent to play Bleeding Edge using a Generative Pretrained Causal Transformer and measure its behavior. Third, we project human and AI gameplay to the proposed behavior manifold to compare and contrast. This allows us to interpret differences in policy as higher-level behavioral concepts, e.g., we find that while human players exhibit variability in fight-flight and explore-exploit behavior, AI players tend towards uniformity. Furthermore, AI agents predominantly engage in solo play, while humans often engage in cooperative and competitive multi-agent patterns. These stark differences underscore the need for interpretable evaluation, design, and integration of AI in human-aligned applications. Our study advances the alignment discussion in AI and especially generative AI research, offering a measurable framework for interpretable human-agent alignment in multiplayer gaming.

cs.AI↗

arXiv2022

Content Addressable Memory Without Catastrophic Forgetting by Heteroassociation with a Fixed Scaffold

Content-addressable memory (CAM) networks, so-called because stored items can be recalled by partial or corrupted versions of the items, exhibit near-perfect recall of a small number of information-dense patterns below capacity and a 'memory cliff' beyond, such that inserting a single additional pattern results in catastrophic loss of all stored patterns. We propose a novel CAM architecture, Memory Scaffold with Heteroassociation (MESH), that factorizes the problems of internal attractor dynamics and association with external content to generate a CAM continuum without a memory cliff: Small numbers of patterns are stored with complete information recovery matching standard CAMs, while inserting more patterns still results in partial recall of every pattern, with a graceful trade-off between pattern number and pattern richness. Motivated by the architecture of the Entorhinal-Hippocampal memory circuit in the brain, MESH is a tripartite architecture with pairwise interactions that uses a predetermined set of internally stabilized states together with heteroassociation between the internal states and arbitrary external patterns. We show analytically and experimentally that for any number of stored patterns, MESH nearly saturates the total information bound (given by the number of synapses) for CAM networks, outperforming all existing CAM models.

cs.AI↗

arXiv2022

Map Induction: Compositional spatial submap learning for efficient exploration in novel environments

Humans are expert explorers. Understanding the computational cognitive mechanisms that support this efficiency can advance the study of the human mind and enable more efficient exploration algorithms. We hypothesize that humans explore new environments efficiently by inferring the structure of unobserved spaces using spatial information collected from previously explored spaces. This cognitive process can be modeled computationally using program induction in a Hierarchical Bayesian framework that explicitly reasons about uncertainty with strong spatial priors. Using a new behavioral Map Induction Task, we demonstrate that this computational framework explains human exploration behavior better than non-inductive models and outperforms state-of-the-art planning algorithms when applied to a realistic spatial navigation domain.

cs.LG↗